Exercise capacity in diabetes mellitus is predicted by activity status and cardiac size rather than cardiac function: a case control study

Timothy J. Roberts, Andrew T. Burns, Richard J. MacIsaac, Andrew I. MacIsaac, David L. Prior, André La Gerche

The reasons for reduced exercise capacity in diabetes mellitus (DM) remains incompletely understood, although diastolic dysfunction and diabetic cardiomyopathy are often favored explanations. However, there is a paucity of literature detailing cardiac function and reserve during incremental exercise to evaluate its significance and contribution. We sought to determine associations between comprehensive measures of cardiac function during exercise and maximal oxygen consumption ([Formula: see text]peak), with the hypothesis that the reduction in exercise capacity and cardiac function would be associated with co-morbidities and sedentary behavior rather than diabetes itself. This case-control study involved 60 subjects [20 with type 1 DM (T1DM), 20 T2DM, and 10 healthy controls age/sex-matched to each diabetes subtype] performing cardiopulmonary exercise testing and bicycle ergometer echocardiography studies. Measures of biventricular function were assessed during incremental exercise to maximal intensity. T2DM subjects were middle-aged (52 ± 11 years) with a mean T2DM diagnosis of 12 ± 7 years and modest glycemic control (HbA1c 57 ± 12 mmol/mol). T1DM participants were younger (35 ± 8 years), with a 19 ± 10 year history of T1DM and suboptimal glycemic control (HbA1c 65 ± 16 mmol/mol). Participants with T2DM were heavier than their controls (body mass index 29.3 ± 3.4 kg/m2 vs. 24.7 ± 2.9, P = 0.001), performed less exercise (10 ± 12 vs. 28 ± 30 MET hours/week, P = 0.031) and had lower exercise capacity ([Formula: see text]peak = 26 ± 6 vs. 38 ± 8 ml/min/kg, P < 0.0001). These differences were not associated with biventricular systolic or left ventricular (LV) diastolic dysfunction at rest or during exercise. There was no difference in weight, exercise participation or [Formula: see text]peak in T1DM subjects as compared to their controls. After accounting for age, sex and body surface area in a multivariate analysis, significant positive predictors of [Formula: see text]peak were cardiac size (LV end-diastolic volume, LVEDV) and estimated MET-hours, while T2DM was a negative predictor. These combined factors accounted for 80% of the variance in [Formula: see text]peak (P < 0.0001). Exercise capacity is reduced in T2DM subjects relative to matched controls, whereas exercise capacity is preserved in T1DM. There was no evidence of sub-clinical cardiac dysfunction but, rather, there was an association between impaired exercise capacity, small LV volumes and sedentary behavior.